Investor | Family Office.
Studying wonderful businesses, market history, and capital allocation.Sharing my investment memos and research.(No financial advice)
I've seen a lot of comments on Salesforce $CRM and plenty of confusion over why the stock is down despite a double beat, so let's get a few points straight and correct the math:
First, full-year revenue growth without the Informatica acquisition is around 8%. With it, it’s 11% (Informatica added $1.1 billion in ARR). Because the guidance relies on Informatica to keep growth in the double digits, Wall Street is rightfully worried about the organic core business—especially with the fear around AI replacing the traditional "seat-based" licensing model.
Second, the massive EPS growth is heavily driven by the $25 billion share buyback. That is financial engineering, not direct core business profit growth.
However, we need to put Salesforce's guidance of roughly $46 billion in total revenue into perspective. At an 11% growth rate, they are adding roughly $4.5 billion in real USD terms to the top line this year. At that massive scale, almost no pure-play software company is close to adding that much absolute cash. Take ServiceNow, for example: they are growing at around 19% to 22% on a ~$15.7 billion revenue base. That means they are adding roughly $3 billion in new revenue, yet they are trading at close to a 70x earnings multiple!
Marc Benioff is taking a massive bet on his own stock's current valuation. If he is right, buying back $25 billion worth of shares will create enormous shareholder value. If he is wrong, taking on that debt will just add additional pressure on Free Cash Flow and earnings margins.
The reason I think their core growth is slowing down comes down to two things:
1Overall global macro uncertainty: Enterprise budgets are still tight.
2The AI Freeze: Most companies are cutting back on traditional IT spend to figure out their AI strategy, but they aren't actually buying the AI yet. It is a waiting game.
Overall, it is still a massive cash cow. It was a good quarter financially, with Total Remaining Performance Obligation (RPO) sitting at a staggering $67.9 billion.
AnalystBook brings filings, earnings, risks, financials, and business context into one industry-aware research workspace, helping investors develop their own view of the companies they follow.
Follow AnalystBook for a better understanding of public companies.
Anyone who has spent time analysing real businesses knows this. Customers do not simply adopt new technology because it is powerful. They prioritise security, safety, reliability, compliance, and operational control.
We saw this with cloud many years ago. Cloud was clearly useful, but large enterprises did not move everything immediately. Many waited for local data centres, stronger security controls, regional compliance, and better reliability. Even today, many critical operations are still on-premise, hybrid, or kept inside controlled environments.
Now the market seems to believe that many enterprises will suddenly trust either new AI startups or in-house AI applications with their most sensitive data. They are expected to let these systems access customer records, employee workflows, legal documents, financial systems, and engineering work, and then allow them to replace large parts of established enterprise software.
I think that misses how businesses actually operate. Especially in the US, Europe, and other regulated markets, there are contracts, legal reviews, data protection rules, GDPR, internal security teams, procurement processes, compliance requirements, and reputational risk. Even a simple contract with a large established company can take months to approve. Yet somehow we are told these same companies will quickly hand over their core systems to new AI tools.
First, no serious C-level executive wants to take responsibility for giving a raw AI system direct access to sensitive company data without strong controls. Customer data, employee records, sales workflows, legal documents, and engineering work are not playgrounds for experimentation.
Second, most companies will not build everything in-house. The integration work is too hard. The workflows are complex. The agentic layer, token economics, security rules, domain expertise, monitoring, and governance all require real infrastructure. A company may build some internal tools, but most will still rely on established platforms.
Third, what we saw over the weekend is that governments can restrict access to advanced AI models, which makes enterprises cautious about building mission-critical systems around tools that may suddenly face regulatory, availability, or compliance problems.
Fourth, frontier models are powerful, but they are not complete enterprise products by themselves. They are raw intelligence. To become useful inside a business, they need rules, permissions, workflows, data context, audit trails, integrations, and clear use cases.
A model can be extremely powerful, but if it does not understand the customer’s workflow, permissions, business logic, compliance rules, and data structure, it cannot safely complete real work. It may answer questions, but answering questions is not the same as running a business process. This is why I think many frontier models may end up becoming part of enterprise ecosystems rather than replacing them entirely.
Salesforce, Adobe, ServiceNow, Microsoft, Oracle, SAP, HubSpot, and others already sit inside business workflows. They already have customer relationships, security reviews, procurement approval, compliance structures, data models, and distribution. If frontier models want to become useful inside enterprises, the most realistic path is often through these existing platforms.
Salesforce is a good example. Salesforce does not need to own every model. It needs to understand which models are best for which tasks, how to manage tokens efficiently, how to protect customer data, and how to build agents that solve real business problems inside sales, service, marketing, and operations. That is where the new economy may be created. Not simply by replacing software, but by making existing workflows more efficient, reducing costs, automating repetitive work, and helping customers justify the software they already use.
Adobe is an important cautionary tale for three reasons:
1. You are likely underestimating how cheap things can get. Fundamentals. Will. Not. Save. You.
2. Reflexivity is real. Adobe has a permanent talent problem now.
3. You can’t fix a terminal value debate with buybacks
$SPCX $ADBE
The day Adobe was getting sold off, SpaceX popped 19% on its first day of trading. Based on their IPO numbers, SpaceX did about $18.7 billion in revenue for 2025 and posted a $4.9 billion net loss. Now, you never want to bet against Elon Musk.
That is definitely true. SpaceX is most likely one of the highest-barrier-to-entry businesses in the world, maybe the highest, and Elon Musk is the one who can turn this business into a profit machine as he has done in the past.
However, look at the other side of the trade. Adobe will be doing $25+ billion in revenue, with 40%+ free cash flow margins and 35%+ operating income margins. Their revenue is still growing double digits. Their ARR is growing. Their profit is growing.
The issue?
Investors are betting that AI is going to completely take Adobe down. Based on that assumption, the market has beaten its valuation down to an 11.9 P/E ratio and roughly 3.3x sales, ignoring the fact that Adobe has around 1 billion users and is used by around 90% of the Fortune 100.
What happened on June 12th is investors paid SpaceX 93 times revenue on unknown margins in the future, while they sold Adobe at 3 times revenue on 40%+ free cash flow and 35%+ operating margins. One group bought the future assumption that SpaceX will be making $40 to $50 billion in profit, maybe even more for many years to come. The other group sold Adobe on the fear that AI is going to kill it, dumping it at only 10 times profits.
Time will tell which group of investors gets the better return.
The confusion, in my view, is that Adobe is not a newspaper company. Adobe has something very different: data, workflow, technical ability, distribution, domain expertise, and deep relationships with creative professionals and enterprises.
It also has partnerships with frontier AI models, but the important part is not only the model. The important part is where the model is used. Adobe can bring those models directly into the workflows where designers, marketers, agencies, and enterprises already work every day.
For enterprises, the question is not just: “Can AI generate an image?” Of course it can.
The real question is: can the company use that image safely, with proper rights, brand consistency, approvals, audit trails, and strict internal rules?
Most companies will not want random AI outputs sitting outside their workflow, especially if they are using the content commercially. They need control. They need ownership. They need legal clarity. They need workflow.
That is why I do not think the newspaper comparison works here.
I think this is wrong for a number of reasons. While the technology is powerful, the idea that a company can just replace their CRM with an AI ignores how businesses actually operate.
* **AI is useless without structure and good data.**
As we’ve seen so far, AI cannot do the work without strict workflows and clean data. Most companies have terrible data to begin with, so a raw AI is going to be of no use to them.
* **AI doesn't know your business rules.** It doesn't natively understand your workflows, corporate governance, or legal constraints. You cannot simply hand over years of complex business logic for an AI to manage on its own.
* **Incumbents are already delivering.** Enterprise companies aren't waiting around; they are already enabling AI right inside their platforms. They are actively solving customer support, sales, and issue resolution with clear, proven use cases.
* **The trust factor.** In the near term, no company is going to trust a raw AI with their sensitive data. They rely on trusted companies like Salesforce to protect them, because hooking an AI up directly to a database can do serious harm if it hallucinates or exposes the wrong info.
* **Work happens in Slack, not ChatGPT.** Employees don't just log into ChatGPT to do their jobs. Work requires multiple layers of collaboration, which is why platforms like Slack are already becoming the main interface for interacting with AI and teammates at the same time.
* **Intelligence will be commoditized.** Just because AI is smart doesn't mean companies are going to build their own support and CRM systems from scratch. First, they lack the domain experience. Second, token costs are expensive to manage. Third, they won't know which model (ChatGPT, Claude, or open-source) is best for a specific use case. Platforms like Salesforce will handle that routing for them.
I could go on and on. I also don't think sales teams will shrink; instead, I think they will grow as AI makes them faster.
If Salesforce navigates this successfully, they will simply unlock a new line of revenue: **selling labor** (AI agents) instead of just software seats.
Yes, AI will sit on top of software, but the enterprise incumbents who already have the trust and relationships within these companies will be the ones delivering those services.
I know this is a contrarian opinion, especially now that Berkshire $BRK.A is taking part in Alphabet’s private placement while $GOOGL is trading near a very high share price.
Berkshire is no longer exactly the same company it was under Buffett, and maybe this is part of adapting to the post-Buffett era.
Google is still a great business. It is highly cash generative, but I also think the risks are higher than many investors realise.
The first risk is search. For many years, Google benefited from a simple and powerful internet structure: people searched, websites competed through SEO, and advertisers followed the traffic. AI changes that. If answers increasingly happen inside AI tools or AI-powered search experiences, the old search economics may become less monopolistic over time.
The second risk is capital intensity. AI is becoming one of the most capital-intensive investment cycles we have seen in technology. Alphabet’s $80 billion equity raise shows how large this investment cycle has become. Even for a company with Google’s cash generation, this creates dilution risk for existing shareholders and could pressure margins and free cash flow if returns do not come through fast enough.
So yes, Google is still a wonderful business. But wonderful businesses can also become less wonderful if the economics of the core business change and the new growth engine requires far more capital than the old one.
I might be wrong, and Google may execute very well. But from my perspective, the upside does not look as obvious as the market seems to believe when compared with the risks: search disruption, AI capex, dilution, and lower future returns on capital.
When investors value big enterprise software companies like $CRM, $NOW, $ADBE or even $HUBS, I think they sometimes overlook one important asset:
Distribution!!
These companies already sit inside the enterprise. They have customer relationships, partner networks, integrations, procurement trust, and years of workflow history.
If AI intelligence becomes more commoditised over time, distribution may become even more valuable, not less.
A new AI start-up may build a better product, but without the channel, trust, and enterprise access, it is very hard to scale alone.
Many of them may eventually become features, partners, or acquisition targets for the platforms that already own the customer relationship.
Not investment advice. Just how I think about the enterprise AI layer.
I've seen a lot of comments on Salesforce $CRM and plenty of confusion over why the stock is down despite a double beat, so let's get a few points straight and correct the math:
First, full-year revenue growth without the Informatica acquisition is around 8%. With it, it’s 11% (Informatica added $1.1 billion in ARR). Because the guidance relies on Informatica to keep growth in the double digits, Wall Street is rightfully worried about the organic core business—especially with the fear around AI replacing the traditional "seat-based" licensing model.
Second, the massive EPS growth is heavily driven by the $25 billion share buyback. That is financial engineering, not direct core business profit growth.
However, we need to put Salesforce's guidance of roughly $46 billion in total revenue into perspective. At an 11% growth rate, they are adding roughly $4.5 billion in real USD terms to the top line this year. At that massive scale, almost no pure-play software company is close to adding that much absolute cash. Take ServiceNow, for example: they are growing at around 19% to 22% on a ~$15.7 billion revenue base. That means they are adding roughly $3 billion in new revenue, yet they are trading at close to a 70x earnings multiple!
Marc Benioff is taking a massive bet on his own stock's current valuation. If he is right, buying back $25 billion worth of shares will create enormous shareholder value. If he is wrong, taking on that debt will just add additional pressure on Free Cash Flow and earnings margins.
The reason I think their core growth is slowing down comes down to two things:
1Overall global macro uncertainty: Enterprise budgets are still tight.
2The AI Freeze: Most companies are cutting back on traditional IT spend to figure out their AI strategy, but they aren't actually buying the AI yet. It is a waiting game.
Overall, it is still a massive cash cow. It was a good quarter financially, with Total Remaining Performance Obligation (RPO) sitting at a staggering $67.9 billion.
$ADBE
Adobe is down roughly 70% from its peak.
Has the business value changed as much as the share price?
I recently published a detailed deep dive on Adobe. Link is in the comments.
I looked at the products, business quality, bull case, bear case, AI risk, and valuation and tried to answer the main question: is Adobe now cheap, or is the business quality changing?
The valuation part is where it gets interesting for me.
Adobe generated about $9.85 billion of free cash flow in FY2025. For FY2026, my working free cash flow base is around $10.5 billion.
In a stress case, where free cash flow declines and the market gives Adobe a low multiple, I get roughly $200 per share.
In a no-growth case, I get around $240 per share.
In a bear-but-still-durable case, where free cash flow grows slowly, I get around $320 per share.
In a base case, where Adobe keeps growing owner cash at a reasonable rate, I get closer to $480 per share.
The exact number is not the point. I do not treat this as a target price.
The point is that at around a $95–100 billion market cap, the market seems to be pricing Adobe close to a no-growth business.
But if Adobe can keep producing more than $10 billion of free cash flow and grow that cash flow even modestly, the current valuation looks too cheap to me.
Not investment advice, just my opinion.
Please read the full article through the link in the comments 👇😌
No browser required. Our API is the UI. 🔓 Salesforce Headless 360 just exposed our entire platform — apps, workflows, metadata, Agentforce & Slack — as unified APIs, MCP tools & CLI. Build on any surface. Give Agentforce deep, trusted context. Stop just using Salesforce. Start building with it. 🚀
https://t.co/8po9KM75RH
The software bears are dead wrong.
At @Pearson, #Agentforce autonomously handles order statuses, refunds & lost access codes — increasing the percentage of customer questions resolved without any human interaction by 40%, says VP Gabriele Bauman.
At @PenFed, our Agentforce agent handles password resets & account unlocks for employees — reducing total IT tickets by 40%, says EVP & CIO Shree Reddy. This is the Agentic Enterprise delivering real ROI today — not hype. Agenticware, not Software. Read why in the @WSJ by @SebasAHerrera:
https://t.co/3cXBvqt9GY
#Agentforce #AgenticEnterprise #AI
Young Investors listen to this investment advice from Bill Ackman.
Bill Ackman explains that the biggest challenge for young investors is the desire for quick rewards. He emphasizes the importance of a long-term perspective, highlighting how compounding creates massive value over decades rather than overnight.
As a striking example, he notes that Warren Buffett built over 99% of his wealth after the age of 50. Ackman's key advice is to leverage the "extra years" of youth to let investments grow steadily over time.
$baba I wrote this article at the beginning of November, and since then, the stock has dropped close to 30%. Read it here, no paywall
https://t.co/FGD24tEnWw